|
I have a TG501 that I am trying to get repaired for another Tech. The work done so far includes.
Adding R606 - Both voltage rails were out of spec are now good.
U100 - 74L00 Dead IC part of the main 1Mhz ring oscillator.
Tested Crystal
C102
C100
R105
R100
C105
All parts tested good R108 is a little out of spec on the value 338 Ohms. Pulling R108 out of circuit should isolate the ring oscillator and let it start but it is not spinning up. What happens when I turn the power on is pin 6 of U100B sits at about 3.5 V DC and gets stuck. I am sure if I can get the 1Mhz running this plugin will be back to functional but I just can¡¯t get it to kick over. The new IC is a TI part and it will run on a bread board but not in circuit. I will upload a clip of the ring oscillator to the files section. IC Socket has been replaced as well.
Any thoughts on how to kick this in to gear?
|
Zen,
Pardon me if this is a stupid question, but have you checked continuity around U100? I only ask because this is an apparently trivial oscillator, and the only reason I can think of for it not to work is that something is not connected correctly.
Also, are you sure that the crystal is good?
¡ª Jeff Dutky
|
Jeff,
I totally agree with you that this is a trivial oscillator. One of the reasons It has been driving me nuts not being able to get it to start. Getting it to start. There is continuity to all the places that there should be and the components check for value and the caps check ok for leakage. Leakage of the capacitors was checked with a Keithley 2450. I even went so far as to swap in a known good crystal. U100 was confirmed bad but the new one still test good on the breadboard. Pin 6 not quite high 3.6 V which would be a problem for and IC as that is below switching range.
It is driving me nuts I can't find where the 1.5V is going.
Zen
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] < [email protected]> On Behalf Of Jeff Dutky Sent: Sunday, November 28, 2021 11:10 PM To: [email protected]Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running. Zen, Pardon me if this is a stupid question, but have you checked continuity around U100? I only ask because this is an apparently trivial oscillator, and the only reason I can think of for it not to work is that something is not connected correctly. Also, are you sure that the crystal is good? ¡ª Jeff Dutky
|
Zen,
what about continuity where it shouldn't be connected?
-- Jeff Dutky
|
A few years back, I had another Tek TM500 series plugin (I don't remember which one) which used a 74L00 as an oscillator. It also refused to oscillate, even though static measurements indicated the IC looked good. I fixed the problem by swapping the 74L00 with another from inventory. The problem vanished.
|
I keep wondering about the replacement 74L00 too - is it too fast? Are you replacing it with new, and is the newer construction different in performance than original?
What happens if you add capacitance to pin 3 (4 & 5) of U100? A breadboard would have more capacitance than the board, wouldn't it?
Dave
|
An early engineering assignment back in the early 90's was the implementation/migration of a process monitor ring oscillator on a 486 variant (486SL). The product was migrating to the next process node.
HSpice simulation of the ring oscillator showed a significantly smaller output swing - too small. The issue was the new process node was "too fast" and while the oscillator worked, the reduced delay through the chain resulted in the output not having time to transition much. The solution was to add more inverters to the chain. There are other ways to slow the ring down - add more capacitance, use slower transistors (longer L), ...
That your pin 6 is "stuck" at 3.6v makes me wonder if the replacement U100 is "too fast" and the output doesn't have time to transition.
Dave
toggle quoted message
Show quoted text
On Monday, November 29, 2021, 07:23:23 AM PST, Dave Peterson via groups.io <davidpinsf@...> wrote:
I keep wondering about the replacement 74L00 too - is it too fast? Are you replacing it with new, and is the newer construction different in performance than original?
What happens if you add capacitance to pin 3 (4 & 5) of U100? A breadboard would have more capacitance than the board, wouldn't it?
Dave
|
I reasoned that this was a really poor way of building an oscillator and that, since TTL really isn't a good
analog part, circuit operation could be dependent upon the individual part's characteristics. When I build oscillators from logic parts, I usually use Schmitt trigger parts; Tektronix didn't. 74HC132, 74hct132 or 74ls132 might work like a chram. Not wanting to reengineer the Tektronix circuit (and parts of it using the other 2 sections of the part), I took the path of least resistance and dug around in my junk bin of little used ancient parts and came up with another 74L00. The part I used was old - possibly older than Tektronix's part. The circuit sprang to life and that was good enough for me.
Crystal oscillators can be tricky since individual crystal characteristics can vary widely. You might be able to get it to run by playing with the capacitances; however, you might not be able to calibrate the frequency. Try lifting one leg of C105. A 74HC00 might be interesting to try. You can always try an LS00 or an HCT00. Note that you can reduce capacitive loading a bit by lifting one input of each gate and conecting it to +5V.
|
Dealing with the gate inputs is easer said then done. On this one they tied the inputs together at the board level under the IC socket. I am starting to think getting this running will be a redesign of the clock. If I shoot out 1Mhz at pin 8 at a TTL level that should get the unit up and running again. So I might be in to making an adaptor board. That can use the socket to get 5 Vdc and output the 1Mhz. would most likely add to the stability as well with modern parts.
Zen
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] < [email protected]> On Behalf Of SCMenasian Sent: Monday, November 29, 2021 12:29 PM To: [email protected]Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running. I reasoned that this was a really poor way of building an oscillator and that, since TTL really isn't a good analog part, circuit operation could be dependent upon the individual part's characteristics. When I build oscillators from logic parts, I usually use Schmitt trigger parts; Tektronix didn't. 74HC132, 74hct132 or 74ls132 might work like a chram. Not wanting to reengineer the Tektronix circuit (and parts of it using the other 2 sections of the part), I took the path of least resistance and dug around in my junk bin of little used ancient parts and came up with another 74L00. The part I used was old - possibly older than Tektronix's part. The circuit sprang to life and that was good enough for me. Crystal oscillators can be tricky since individual crystal characteristics can vary widely. You might be able to get it to run by playing with the capacitances; however, you might not be able to calibrate the frequency. Try lifting one leg of C105. A 74HC00 might be interesting to try. You can always try an LS00 or an HCT00. Note that you can reduce capacitive loading a bit by lifting one input of each gate and conecting it to +5V.
|
Would a standard 14 pin oscillator package work?? 14 = vcc, 7 = gnd, 8 = output, and no other pins are connected.? A 14 pin wirewrap socket with pins removed (and an under socket jumper on the socket) would save having to make a board.
I think that 1 Mhz is available.?? you could do something similar with a half-dip size and a few more jumpers.
Harvey
toggle quoted message
Show quoted text
On 11/29/2021 1:54 PM, Zentronics42@... wrote: Dealing with the gate inputs is easer said then done. On this one they tied the inputs together at the board level under the IC socket. I am starting to think getting this running will be a redesign of the clock. If I shoot out 1Mhz at pin 8 at a TTL level that should get the unit up and running again. So I might be in to making an adaptor board. That can use the socket to get 5 Vdc and output the 1Mhz. would most likely add to the stability as well with modern parts.
Zen
-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of SCMenasian
Sent: Monday, November 29, 2021 12:29 PM
To: [email protected]
Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running.
I reasoned that this was a really poor way of building an oscillator and that, since TTL really isn't a good analog part, circuit operation could be dependent upon the individual part's characteristics. When I build oscillators from logic parts, I usually use Schmitt trigger parts; Tektronix didn't. 74HC132, 74hct132 or 74ls132 might work like a chram. Not wanting to reengineer the Tektronix circuit (and parts of it using the other 2 sections of the part), I took the path of least resistance and dug around in my junk bin of little used ancient parts and came up with another 74L00. The part I used was old - possibly older than Tektronix's part. The circuit sprang to life and that was good enough for me.
Crystal oscillators can be tricky since individual crystal characteristics can vary widely. You might be able to get it to run by playing with the capacitances; however, you might not be able to calibrate the frequency. Try lifting one leg of C105. A 74HC00 might be interesting to try. You can always try an LS00 or an HCT00. Note that you can reduce capacitive loading a bit by lifting one input of each gate and conecting it to +5V.
|
It should I would like to keep the adjustment capability for future calibrations as this unit will need to travel 1/2 way around the world when it goes back to its owner. Unless the 1Mhz clocks are good enough at this point that they don¡¯t need adjustment?
That would be wildly convenient if it would work the socket uses standard TTL with 7 gnd, 14 VCC = 5.2Vdc. And in this case 8 is the output if the ring. As long as the frequency stability is there this might be a solution to the issue, And great to know going forward.
Zen
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] < [email protected]> On Behalf Of Harvey White Sent: Monday, November 29, 2021 2:10 PM To: [email protected]Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running. Would a standard 14 pin oscillator package work? 14 = vcc, 7 = gnd, 8 = output, and no other pins are connected. A 14 pin wirewrap socket with pins removed (and an under socket jumper on the socket) would save having to make a board. I think that 1 Mhz is available. you could do something similar with a half-dip size and a few more jumpers. Harvey On 11/29/2021 1:54 PM, Zentronics42@... wrote: Dealing with the gate inputs is easer said then done. On this one they tied the inputs together at the board level under the IC socket. I am starting to think getting this running will be a redesign of the clock. If I shoot out 1Mhz at pin 8 at a TTL level that should get the unit up and running again. So I might be in to making an adaptor board. That can use the socket to get 5 Vdc and output the 1Mhz. would most likely add to the stability as well with modern parts.
Zen
-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of
SCMenasian
Sent: Monday, November 29, 2021 12:29 PM
To: [email protected]
Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running.
I reasoned that this was a really poor way of building an oscillator and that, since TTL really isn't a good analog part, circuit operation could be dependent upon the individual part's characteristics. When I build oscillators from logic parts, I usually use Schmitt trigger parts; Tektronix didn't. 74HC132, 74hct132 or 74ls132 might work like a chram. Not wanting to reengineer the Tektronix circuit (and parts of it using the other 2 sections of the part), I took the path of least resistance and dug around in my junk bin of little used ancient parts and came up with another 74L00. The part I used was old - possibly older than Tektronix's part. The circuit sprang to life and that was good enough for me.
Crystal oscillators can be tricky since individual crystal characteristics can vary widely. You might be able to get it to run by playing with the capacitances; however, you might not be able to calibrate the frequency. Try lifting one leg of C105. A 74HC00 might be interesting to try. You can always try an LS00 or an HCT00. Note that you can reduce capacitive loading a bit by lifting one input of each gate and conecting it to +5V.
|
Zen,
I converted all of my TG501s over to "Option 1" Units with the 5Mhz Oscillator. This might be a solution as well.
--
Michael Lynch
Dardanelle, AR
|
On Monday 29 November 2021 02:09:55 pm Harvey White wrote: Would a standard 14 pin oscillator package work?? 14 = vcc, 7 = gnd, 8 =
output, and no other pins are connected.? A 14 pin wirewrap socket with
pins removed (and an under socket jumper on the socket) would save
having to make a board.
I think that 1 Mhz is available.?? you could do something similar with a
half-dip size and a few more jumpers. I happen to have a whole pile of those, if anybody wants one. Feel free to contact me off-list... Harvey
On 11/29/2021 1:54 PM, Zentronics42@... wrote:
Dealing with the gate inputs is easer said then done. On this one they tied the inputs together at the board level under the IC socket. I am starting to think getting this running will be a redesign of the clock. If I shoot out 1Mhz at pin 8 at a TTL level that should get the unit up and running again. So I might be in to making an adaptor board. That can use the socket to get 5 Vdc and output the 1Mhz. would most likely add to the stability as well with modern parts.
Zen
-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of SCMenasian
Sent: Monday, November 29, 2021 12:29 PM
To: [email protected]
Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running.
I reasoned that this was a really poor way of building an oscillator and that, since TTL really isn't a good analog part, circuit operation could be dependent upon the individual part's characteristics. When I build oscillators from logic parts, I usually use Schmitt trigger parts; Tektronix didn't. 74HC132, 74hct132 or 74ls132 might work like a chram. Not wanting to reengineer the Tektronix circuit (and parts of it using the other 2 sections of the part), I took the path of least resistance and dug around in my junk bin of little used ancient parts and came up with another 74L00. The part I used was old - possibly older than Tektronix's part. The circuit sprang to life and that was good enough for me.
Crystal oscillators can be tricky since individual crystal characteristics can vary widely. You might be able to get it to run by playing with the capacitances; however, you might not be able to calibrate the frequency. Try lifting one leg of C105. A 74HC00 might be interesting to try. You can always try an LS00 or an HCT00. Note that you can reduce capacitive loading a bit by lifting one input of each gate and conecting it to +5V. -- Member of the toughest, meanest, deadliest, most unrelenting -- and ablest -- form of life in this section of space, ?a critter that can be killed but can't be tamed. ?--Robert A. Heinlein, "The Puppet Masters" - Information is more dangerous than cannon to a society ruled by lies. --James M Dakin
|
On Mon, Nov 29, 2021 at 11:39 AM, Michael W. Lynch wrote:
Zen,
I converted all of my TG501s over to "Option 1" Units with the 5Mhz
Oscillator. This might be a solution as well.
--
Michael Lynch
Dardanelle, AR
Michael, I'd like to consider making this conversion. My TG501 timing is doing fine with its 1MHz oscillator, but my finicky nature would like to achieve the +/- 5x10^-7 option 1 seems to give. I would think this would directly bear on the accuracy of the 1ns timing marks? Do you have something of a BOM for the conversion? Feel free to reply off list if you prefer. Thanks, Dave
|
Dave,
I had thought that I had seen a video by either Jared Cabot (NFM) or Zen (Zenwizard) showing the addition of Option 01 to a TG 501, but I can't seem to find it now. I may have been thinking of Jared's video about adding a similar option to a DC503A.
The TG 501 Option 01 seems to consist of basically two components: U50 (a 7490 TTL decade counter) and Y50 (a 5MHz crystal oscillator, possibly ovenized?). The part numbers for Y50 are either Electronic Research Co. EROS-600-TK-9 (SN below B034510) or Bliley Electric Co. T15B141 (SN B034510 and above). I imagine that any 5 MHz TXCO or OCXO can be made to work, with effort.
The service manual has this to say about Option 1:
The TG 501 Option 01 is equipped with a temperature-compensated, 5 MHz crystal oscillator for a highly stable and precise internal clock. U50 provides countdown circuitry for the 5 MHz crystal.
Remove U100 and disconnect pin 3 of U350 before using external clocks. To disconnect pin 3 remove U350, bend pin 3 out then insert U350 back in tcurciut board.
The main circuit board jumpers that must be installed for Option 01 are: J3 and J7 (see. Fig. 5-2 for jumper location).
-- Jeff Dutky
|
Thanks Jeff,
Michael pointed me to the appropriate section of the SM (RTFM Dave). I see there is an oscillator available on eBay. Not sure I feel the need to spend another $50 on eBay just today. I just bought an S-53 to enable viewing 1GHz signals. It was an impulse buy, sure, but this has been on my mind. And now I feel the need to have that capability. So I need to pace my spending. (Why start now!?)
Another intriguing possibility I've found in the SM is the possibility of using an external clock. So it seems I could pipe the Leo Bodnar GPSDO directly into the TG501. Now there's some time mark stability and accuracy! It should be easy to add a BNC to the back of a TM50x.
BTW, my homemade FR4 trimmer has set up nicely. I'm going to wait another day before trying it out on C515. I am looking forward to using it on the 2ns and 1ns Filter caps. I feel I've already tuned them well, but this all seems to require repeated fine tuning. And hopefully with the S-53 I'll be able to observe the 1ns signal at full BW (this will go in a 7S12 with S-6 sampler which is working quite well with my now working S-52).
Dave
toggle quoted message
Show quoted text
On Friday, December 3, 2021, 03:21:58 PM PST, Jeff Dutky <jeff.dutky@...> wrote:
Dave,
I had thought that I had seen a video by either Jared Cabot (NFM) or Zen (Zenwizard) showing the addition of Option 01 to a TG 501, but I can't seem to find it now. I may have been thinking of Jared's video about adding a similar option to a DC503A.
The TG 501 Option 01 seems to consist of basically two components: U50 (a 7490 TTL decade counter) and Y50 (a 5MHz crystal oscillator, possibly ovenized?). The part numbers for Y50 are either Electronic Research Co. EROS-600-TK-9 (SN below B034510) or Bliley Electric Co. T15B141 (SN B034510 and above). I imagine that any 5 MHz TXCO or OCXO can be made to work, with effort.
The service manual has this to say about Option 1:
The TG 501 Option 01 is equipped with a temperature-compensated, 5 MHz crystal oscillator for a highly stable and precise internal clock. U50 provides countdown circuitry for the 5 MHz crystal.
Remove U100 and disconnect pin 3 of U350 before using external clocks. To disconnect pin 3 remove U350, bend pin 3 out then insert U350 back in tcurciut board.
The main circuit board jumpers that must be installed for Option 01 are: J3 and J7 (see. Fig. 5-2 for jumper location).
-- Jeff Dutky
|
I was going to suggest Leo¡¯s GPSDO but you figured that out before I got ¡®round to typing. I know that the desire for high precision/high accuracy can be an obsession! For calibrating scopes, something less elaborate more than suffices. Even the cheapest CPU clock oscillators are good to 200ppm. You¡¯ll never see that error on a scope. But high accuracy clocks are now available for a song. I have a rubidium standard, and although I don¡¯t do any work requiring 10ppb accuracy, the fact that I have one caused me to check every frequency counter and every frequency source I own and tweak out whatever errors could be tweaked out. Ridiculous! But fun.
Enjoy the Quest For the Last Digit!
¡ª Cheers
Tom
Sent from my iThing, so please forgive typos and brevity.
toggle quoted message
Show quoted text
On Dec 3, 2021, at 3:35 PM, Dave Peterson via groups.io <davidpinsf@...> wrote:
? Thanks Jeff,
Michael pointed me to the appropriate section of the SM (RTFM Dave). I see there is an oscillator available on eBay. Not sure I feel the need to spend another $50 on eBay just today. I just bought an S-53 to enable viewing 1GHz signals. It was an impulse buy, sure, but this has been on my mind. And now I feel the need to have that capability. So I need to pace my spending. (Why start now!?)
Another intriguing possibility I've found in the SM is the possibility of using an external clock. So it seems I could pipe the Leo Bodnar GPSDO directly into the TG501. Now there's some time mark stability and accuracy! It should be easy to add a BNC to the back of a TM50x.
BTW, my homemade FR4 trimmer has set up nicely. I'm going to wait another day before trying it out on C515. I am looking forward to using it on the 2ns and 1ns Filter caps. I feel I've already tuned them well, but this all seems to require repeated fine tuning. And hopefully with the S-53 I'll be able to observe the 1ns signal at full BW (this will go in a 7S12 with S-6 sampler which is working quite well with my now working S-52).
Dave
On Friday, December 3, 2021, 03:21:58 PM PST, Jeff Dutky <jeff.dutky@...> wrote:
Dave,
I had thought that I had seen a video by either Jared Cabot (NFM) or Zen (Zenwizard) showing the addition of Option 01 to a TG 501, but I can't seem to find it now. I may have been thinking of Jared's video about adding a similar option to a DC503A.
The TG 501 Option 01 seems to consist of basically two components: U50 (a 7490 TTL decade counter) and Y50 (a 5MHz crystal oscillator, possibly ovenized?). The part numbers for Y50 are either Electronic Research Co. EROS-600-TK-9 (SN below B034510) or Bliley Electric Co. T15B141 (SN B034510 and above). I imagine that any 5 MHz TXCO or OCXO can be made to work, with effort.
The service manual has this to say about Option 1:
The TG 501 Option 01 is equipped with a temperature-compensated, 5 MHz crystal oscillator for a highly stable and precise internal clock. U50 provides countdown circuitry for the 5 MHz crystal.
Remove U100 and disconnect pin 3 of U350 before using external clocks. To disconnect pin 3 remove U350, bend pin 3 out then insert U350 back in tcurciut board.
The main circuit board jumpers that must be installed for Option 01 are: J3 and J7 (see. Fig. 5-2 for jumper location).
-- Jeff Dutky
|
On Sat, Dec 4, 2021 at 08:34 AM, Dave Peterson wrote:
BTW, my homemade FR4 trimmer has set up nicely. I'm going to wait another day
before trying it out on C515. I am looking forward to using it on the 2ns and
1ns Filter caps. I feel I've already tuned them well, but this all seems to
require repeated fine tuning.
On my TG501A I went through the adjustment process about 4 times before I had it nailed right down. A number of the adjustments interact with each other so a couple passes will get everything playing nicely with each other.
|
This will be an extended troubleshooting experience. However one I asked for when I offered to help. I am running in to several marginal ICs in this unit. So I have a few IC testers on the way, I think I am going to need them. I am currently focused on U330, U300, and U235 in the switch circuit. With an external clock I can get the unit to behave enough to get good time marks out of it. However the frequency is no where near correct. And only 3 positions of the dial even look correct. All the other positions of the dial result in double time marks. I can post a picture if it will make more sense. I will also have a part 1 video out soon. The time marks look normal at 2us 1us and .5us everything else is doubled up and has time marks on top of slower time marks.
Variable timing seems to work as it should. And with an external clock now gives me readouts instead of 0.4 and not moving. So making some headway but still a ways to go before I get this back up to running.
Zen
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] < [email protected]> On Behalf Of Jared Cabot via groups.io Sent: Saturday, December 4, 2021 2:03 AM To: [email protected]Subject: Re: [TekScopes] Sick TG501 Main 1Mhz ref not running. On Sat, Dec 4, 2021 at 08:34 AM, Dave Peterson wrote:
BTW, my homemade FR4 trimmer has set up nicely. I'm going to wait
another day before trying it out on C515. I am looking forward to
using it on the 2ns and 1ns Filter caps. I feel I've already tuned
them well, but this all seems to require repeated fine tuning.
On my TG501A I went through the adjustment process about 4 times before I had it nailed right down. A number of the adjustments interact with each other so a couple passes will get everything playing nicely with each other.
|
Zen,
I had a similar problem (double time marks) on a TG501 unit several months ago. It turned out that one finger of the drum switch assembly was not opening when it should have due to mechanical damage. I found the problem using the switch matrix in the SM which allowed me to figure out the proper position of each switch finger. The matrix is located on the ¡°Display and Switch Details¡± page of the schematic section in the SM.
On the subject of vintage IC testing: The Tauntek tester lets you see pin voltages and overall current for a given chip as well as the usual logic testing.
Shaun M
|
Dave Peterson
Michael W. Lynch
Jared Cabot
Shaun M